]> git.saurik.com Git - wxWidgets.git/blame - src/common/string.cpp
Replaced (char*)wxFNCONV with (const char*)wxFNCONV
[wxWidgets.git] / src / common / string.cpp
CommitLineData
c801d85f
KB
1/////////////////////////////////////////////////////////////////////////////
2// Name: string.cpp
3// Purpose: wxString class
4// Author: Vadim Zeitlin
5// Modified by:
6// Created: 29/01/98
7// RCS-ID: $Id$
8// Copyright: (c) 1998 Vadim Zeitlin <zeitlin@dptmaths.ens-cachan.fr>
dd1eaa89 9// Licence: wxWindows license
c801d85f
KB
10/////////////////////////////////////////////////////////////////////////////
11
12#ifdef __GNUG__
30b21f9a 13 #pragma implementation "string.h"
c801d85f
KB
14#endif
15
16/*
17 * About ref counting:
18 * 1) all empty strings use g_strEmpty, nRefs = -1 (set in Init())
19 * 2) AllocBuffer() sets nRefs to 1, Lock() increments it by one
20 * 3) Unlock() decrements nRefs and frees memory if it goes to 0
21 */
22
23// ===========================================================================
24// headers, declarations, constants
25// ===========================================================================
26
27// For compilers that support precompilation, includes "wx.h".
28#include "wx/wxprec.h"
29
30#ifdef __BORLANDC__
30b21f9a 31 #pragma hdrstop
c801d85f
KB
32#endif
33
34#ifndef WX_PRECOMP
3c024cc2
VZ
35 #include "wx/defs.h"
36 #include "wx/string.h"
37 #include "wx/intl.h"
6b769f3d
OK
38#if wxUSE_THREADS
39 #include <wx/thread.h>
40#endif
c801d85f
KB
41#endif
42
43#include <ctype.h>
44#include <string.h>
45#include <stdlib.h>
46
ce3ed50d 47#ifdef __SALFORDC__
30b21f9a 48 #include <clib.h>
ce3ed50d
JS
49#endif
50
ede25f5b 51#if wxUSE_WCSRTOMBS
fb4e5803
VZ
52 #include <wchar.h> // for wcsrtombs(), see comments where it's used
53#endif // GNU
54
c801d85f
KB
55#ifdef WXSTRING_IS_WXOBJECT
56 IMPLEMENT_DYNAMIC_CLASS(wxString, wxObject)
57#endif //WXSTRING_IS_WXOBJECT
58
ec2ba3aa
OK
59#if wxUSE_UNICODE
60#undef wxUSE_EXPERIMENTAL_PRINTF
61#define wxUSE_EXPERIMENTAL_PRINTF 1
62#endif
63
3168a13f
VZ
64// allocating extra space for each string consumes more memory but speeds up
65// the concatenation operations (nLen is the current string's length)
77ca46e7
VZ
66// NB: EXTRA_ALLOC must be >= 0!
67#define EXTRA_ALLOC (19 - nLen % 16)
3168a13f 68
c801d85f
KB
69// ---------------------------------------------------------------------------
70// static class variables definition
71// ---------------------------------------------------------------------------
72
8de2e39c 73#ifdef wxSTD_STRING_COMPATIBILITY
566b84d2 74 const size_t wxString::npos = wxSTRING_MAXLEN;
8de2e39c 75#endif // wxSTD_STRING_COMPATIBILITY
c801d85f 76
3168a13f
VZ
77// ----------------------------------------------------------------------------
78// static data
79// ----------------------------------------------------------------------------
c801d85f 80
3c024cc2
VZ
81// for an empty string, GetStringData() will return this address: this
82// structure has the same layout as wxStringData and it's data() method will
83// return the empty string (dummy pointer)
84static const struct
85{
86 wxStringData data;
2bb67b80
OK
87 wxChar dummy;
88} g_strEmpty = { {-1, 0, 0}, _T('\0') };
3c024cc2 89
c801d85f 90// empty C style string: points to 'string data' byte of g_strEmpty
2bb67b80 91extern const wxChar WXDLLEXPORT *g_szNul = &g_strEmpty.dummy;
c801d85f 92
89b892a2
VZ
93// ----------------------------------------------------------------------------
94// conditional compilation
95// ----------------------------------------------------------------------------
96
dcf924a3
RR
97#if !defined(__WXSW__) && wxUSE_UNICODE
98 #ifdef wxUSE_EXPERIMENTAL_PRINTF
99 #undef wxUSE_EXPERIMENTAL_PRINTF
100 #endif
101 #define wxUSE_EXPERIMENTAL_PRINTF 1
102#endif
103
89b892a2
VZ
104// we want to find out if the current platform supports vsnprintf()-like
105// function: for Unix this is done with configure, for Windows we test the
106// compiler explicitly.
107#ifdef __WXMSW__
3f4a0c5b 108 #ifdef __VISUALC__
2bb67b80 109 #define wxVsnprintf _vsnprintf
89b892a2
VZ
110 #endif
111#else // !Windows
112 #ifdef HAVE_VSNPRINTF
2bb67b80 113 #define wxVsnprintf vsnprintf
89b892a2
VZ
114 #endif
115#endif // Windows/!Windows
116
2bb67b80 117#ifndef wxVsnprintf
89b892a2
VZ
118 // in this case we'll use vsprintf() (which is ANSI and thus should be
119 // always available), but it's unsafe because it doesn't check for buffer
120 // size - so give a warning
2bb67b80 121 #define wxVsnprintf(buffer,len,format,argptr) vsprintf(buffer,format, argptr)
566b84d2 122
57493f9f
VZ
123 #if defined(__VISUALC__)
124 #pragma message("Using sprintf() because no snprintf()-like function defined")
125 #elif defined(__GNUG__) && !defined(__UNIX__)
126 #warning "Using sprintf() because no snprintf()-like function defined"
127 #elif defined(__MWERKS__)
4e57b0d4 128 #warning "Using sprintf() because no snprintf()-like function defined"
57493f9f 129 #endif //compiler
3f4a0c5b 130#endif // no vsnprintf
89b892a2 131
227b5cd7
VZ
132#ifdef _AIX
133 // AIX has vsnprintf, but there's no prototype in the system headers.
134 extern "C" int vsnprintf(char* str, size_t n, const char* format, va_list ap);
135#endif
136
3168a13f 137// ----------------------------------------------------------------------------
c801d85f 138// global functions
3168a13f 139// ----------------------------------------------------------------------------
c801d85f 140
a533f5c1 141#if defined(wxSTD_STRING_COMPATIBILITY) && wxUSE_STD_IOSTREAM
c801d85f
KB
142
143// MS Visual C++ version 5.0 provides the new STL headers as well as the old
144// iostream ones.
145//
146// ATTN: you can _not_ use both of these in the same program!
a38b83c3 147
3f4a0c5b 148istream& operator>>(istream& is, wxString& WXUNUSED(str))
c801d85f
KB
149{
150#if 0
151 int w = is.width(0);
152 if ( is.ipfx(0) ) {
3f4a0c5b 153 streambuf *sb = is.rdbuf();
c801d85f
KB
154 str.erase();
155 while ( true ) {
156 int ch = sb->sbumpc ();
157 if ( ch == EOF ) {
3f4a0c5b 158 is.setstate(ios::eofbit);
c801d85f
KB
159 break;
160 }
161 else if ( isspace(ch) ) {
162 sb->sungetc();
163 break;
164 }
dd1eaa89 165
c801d85f
KB
166 str += ch;
167 if ( --w == 1 )
168 break;
169 }
170 }
171
172 is.isfx();
173 if ( str.length() == 0 )
3f4a0c5b 174 is.setstate(ios::failbit);
c801d85f
KB
175#endif
176 return is;
177}
178
179#endif //std::string compatibility
180
3168a13f
VZ
181// ----------------------------------------------------------------------------
182// private classes
183// ----------------------------------------------------------------------------
184
185// this small class is used to gather statistics for performance tuning
186//#define WXSTRING_STATISTICS
187#ifdef WXSTRING_STATISTICS
188 class Averager
189 {
190 public:
191 Averager(const char *sz) { m_sz = sz; m_nTotal = m_nCount = 0; }
2c3b684c 192 ~Averager()
3168a13f
VZ
193 { printf("wxString: average %s = %f\n", m_sz, ((float)m_nTotal)/m_nCount); }
194
c86f1403 195 void Add(size_t n) { m_nTotal += n; m_nCount++; }
3168a13f
VZ
196
197 private:
c86f1403 198 size_t m_nCount, m_nTotal;
3168a13f
VZ
199 const char *m_sz;
200 } g_averageLength("allocation size"),
201 g_averageSummandLength("summand length"),
202 g_averageConcatHit("hit probability in concat"),
203 g_averageInitialLength("initial string length");
204
205 #define STATISTICS_ADD(av, val) g_average##av.Add(val)
206#else
207 #define STATISTICS_ADD(av, val)
208#endif // WXSTRING_STATISTICS
209
c801d85f
KB
210// ===========================================================================
211// wxString class core
212// ===========================================================================
213
214// ---------------------------------------------------------------------------
215// construction
216// ---------------------------------------------------------------------------
217
c801d85f 218// constructs string of <nLength> copies of character <ch>
2bb67b80 219wxString::wxString(wxChar ch, size_t nLength)
c801d85f
KB
220{
221 Init();
222
223 if ( nLength > 0 ) {
224 AllocBuffer(nLength);
f1da2f03 225
2bb67b80
OK
226#if wxUSE_UNICODE
227 // memset only works on char
228 for (size_t n=0; n<nLength; n++) m_pchData[n] = ch;
229#else
c801d85f 230 memset(m_pchData, ch, nLength);
2bb67b80 231#endif
c801d85f
KB
232 }
233}
234
235// takes nLength elements of psz starting at nPos
2bb67b80 236void wxString::InitWith(const wxChar *psz, size_t nPos, size_t nLength)
c801d85f
KB
237{
238 Init();
239
2bb67b80 240 wxASSERT( nPos <= wxStrlen(psz) );
c801d85f 241
566b84d2 242 if ( nLength == wxSTRING_MAXLEN )
2bb67b80 243 nLength = wxStrlen(psz + nPos);
c801d85f 244
3168a13f
VZ
245 STATISTICS_ADD(InitialLength, nLength);
246
c801d85f
KB
247 if ( nLength > 0 ) {
248 // trailing '\0' is written in AllocBuffer()
249 AllocBuffer(nLength);
2bb67b80 250 memcpy(m_pchData, psz + nPos, nLength*sizeof(wxChar));
c801d85f
KB
251 }
252}
dd1eaa89 253
8de2e39c 254#ifdef wxSTD_STRING_COMPATIBILITY
c801d85f 255
c801d85f
KB
256// poor man's iterators are "void *" pointers
257wxString::wxString(const void *pStart, const void *pEnd)
258{
2bb67b80
OK
259 InitWith((const wxChar *)pStart, 0,
260 (const wxChar *)pEnd - (const wxChar *)pStart);
c801d85f
KB
261}
262
263#endif //std::string compatibility
264
2bb67b80
OK
265#if wxUSE_UNICODE
266
267// from multibyte string
cf2f341a 268wxString::wxString(const char *psz, wxMBConv& conv, size_t nLength)
2bb67b80
OK
269{
270 // first get necessary size
435595e0 271 size_t nLen = psz ? conv.MB2WC((wchar_t *) NULL, psz, 0) : 0;
2bb67b80
OK
272
273 // nLength is number of *Unicode* characters here!
eea4f86a 274 if ((nLen != (size_t)-1) && (nLen > nLength))
2bb67b80
OK
275 nLen = nLength;
276
277 // empty?
eea4f86a 278 if ( (nLen != 0) && (nLen != (size_t)-1) ) {
2bb67b80
OK
279 AllocBuffer(nLen);
280 conv.MB2WC(m_pchData, psz, nLen);
281 }
282 else {
283 Init();
284 }
285}
286
287#else
288
0f3e3e0c 289#if wxUSE_WCHAR_T
c801d85f
KB
290// from wide string
291wxString::wxString(const wchar_t *pwz)
292{
293 // first get necessary size
435595e0 294 size_t nLen = pwz ? wxWC2MB((char *) NULL, pwz, 0) : 0;
c801d85f
KB
295
296 // empty?
eea4f86a 297 if ( (nLen != 0) && (nLen != (size_t)-1) ) {
c801d85f 298 AllocBuffer(nLen);
2bb67b80 299 wxWC2MB(m_pchData, pwz, nLen);
c801d85f
KB
300 }
301 else {
302 Init();
303 }
304}
0f3e3e0c 305#endif
c801d85f 306
2bb67b80
OK
307#endif
308
c801d85f
KB
309// ---------------------------------------------------------------------------
310// memory allocation
311// ---------------------------------------------------------------------------
312
313// allocates memory needed to store a C string of length nLen
314void wxString::AllocBuffer(size_t nLen)
315{
316 wxASSERT( nLen > 0 ); //
317 wxASSERT( nLen <= INT_MAX-1 ); // max size (enough room for 1 extra)
318
3168a13f
VZ
319 STATISTICS_ADD(Length, nLen);
320
c801d85f
KB
321 // allocate memory:
322 // 1) one extra character for '\0' termination
323 // 2) sizeof(wxStringData) for housekeeping info
3168a13f 324 wxStringData* pData = (wxStringData*)
2bb67b80 325 malloc(sizeof(wxStringData) + (nLen + EXTRA_ALLOC + 1)*sizeof(wxChar));
c801d85f 326 pData->nRefs = 1;
c801d85f 327 pData->nDataLength = nLen;
3168a13f 328 pData->nAllocLength = nLen + EXTRA_ALLOC;
c801d85f 329 m_pchData = pData->data(); // data starts after wxStringData
2bb67b80 330 m_pchData[nLen] = _T('\0');
c801d85f
KB
331}
332
c801d85f
KB
333// must be called before changing this string
334void wxString::CopyBeforeWrite()
335{
336 wxStringData* pData = GetStringData();
337
338 if ( pData->IsShared() ) {
339 pData->Unlock(); // memory not freed because shared
c86f1403 340 size_t nLen = pData->nDataLength;
3168a13f 341 AllocBuffer(nLen);
2bb67b80 342 memcpy(m_pchData, pData->data(), nLen*sizeof(wxChar));
c801d85f
KB
343 }
344
3bbb630a 345 wxASSERT( !GetStringData()->IsShared() ); // we must be the only owner
c801d85f
KB
346}
347
348// must be called before replacing contents of this string
349void wxString::AllocBeforeWrite(size_t nLen)
350{
351 wxASSERT( nLen != 0 ); // doesn't make any sense
352
353 // must not share string and must have enough space
3168a13f 354 wxStringData* pData = GetStringData();
c801d85f
KB
355 if ( pData->IsShared() || (nLen > pData->nAllocLength) ) {
356 // can't work with old buffer, get new one
357 pData->Unlock();
358 AllocBuffer(nLen);
359 }
471aebdd
VZ
360 else {
361 // update the string length
362 pData->nDataLength = nLen;
363 }
c801d85f 364
f1da2f03 365 wxASSERT( !GetStringData()->IsShared() ); // we must be the only owner
c801d85f
KB
366}
367
dd1eaa89 368// allocate enough memory for nLen characters
c86f1403 369void wxString::Alloc(size_t nLen)
dd1eaa89
VZ
370{
371 wxStringData *pData = GetStringData();
372 if ( pData->nAllocLength <= nLen ) {
9fbd8b8d
VZ
373 if ( pData->IsEmpty() ) {
374 nLen += EXTRA_ALLOC;
375
376 wxStringData* pData = (wxStringData*)
2bb67b80 377 malloc(sizeof(wxStringData) + (nLen + 1)*sizeof(wxChar));
9fbd8b8d
VZ
378 pData->nRefs = 1;
379 pData->nDataLength = 0;
380 pData->nAllocLength = nLen;
381 m_pchData = pData->data(); // data starts after wxStringData
2bb67b80 382 m_pchData[0u] = _T('\0');
9fbd8b8d 383 }
3168a13f
VZ
384 else if ( pData->IsShared() ) {
385 pData->Unlock(); // memory not freed because shared
c86f1403 386 size_t nOldLen = pData->nDataLength;
3168a13f 387 AllocBuffer(nLen);
2bb67b80 388 memcpy(m_pchData, pData->data(), nOldLen*sizeof(wxChar));
3168a13f 389 }
dd1eaa89 390 else {
3168a13f
VZ
391 nLen += EXTRA_ALLOC;
392
dd1eaa89 393 wxStringData *p = (wxStringData *)
2bb67b80 394 realloc(pData, sizeof(wxStringData) + (nLen + 1)*sizeof(wxChar));
3168a13f
VZ
395
396 if ( p == NULL ) {
397 // @@@ what to do on memory error?
398 return;
dd1eaa89 399 }
3168a13f
VZ
400
401 // it's not important if the pointer changed or not (the check for this
402 // is not faster than assigning to m_pchData in all cases)
403 p->nAllocLength = nLen;
404 m_pchData = p->data();
dd1eaa89
VZ
405 }
406 }
407 //else: we've already got enough
408}
409
410// shrink to minimal size (releasing extra memory)
411void wxString::Shrink()
412{
413 wxStringData *pData = GetStringData();
3bbb630a
VZ
414
415 // this variable is unused in release build, so avoid the compiler warning by
416 // just not declaring it
417#ifdef __WXDEBUG__
418 void *p =
419#endif
2bb67b80 420 realloc(pData, sizeof(wxStringData) + (pData->nDataLength + 1)*sizeof(wxChar));
3bbb630a 421
3168a13f 422 wxASSERT( p != NULL ); // can't free memory?
dd1eaa89
VZ
423 wxASSERT( p == pData ); // we're decrementing the size - block shouldn't move!
424}
425
c801d85f 426// get the pointer to writable buffer of (at least) nLen bytes
2bb67b80 427wxChar *wxString::GetWriteBuf(size_t nLen)
c801d85f
KB
428{
429 AllocBeforeWrite(nLen);
097c080b
VZ
430
431 wxASSERT( GetStringData()->nRefs == 1 );
432 GetStringData()->Validate(FALSE);
433
c801d85f
KB
434 return m_pchData;
435}
436
097c080b
VZ
437// put string back in a reasonable state after GetWriteBuf
438void wxString::UngetWriteBuf()
439{
2bb67b80 440 GetStringData()->nDataLength = wxStrlen(m_pchData);
097c080b
VZ
441 GetStringData()->Validate(TRUE);
442}
443
c801d85f
KB
444// ---------------------------------------------------------------------------
445// data access
446// ---------------------------------------------------------------------------
447
448// all functions are inline in string.h
449
450// ---------------------------------------------------------------------------
451// assignment operators
452// ---------------------------------------------------------------------------
453
dd1eaa89 454// helper function: does real copy
2bb67b80 455void wxString::AssignCopy(size_t nSrcLen, const wxChar *pszSrcData)
c801d85f
KB
456{
457 if ( nSrcLen == 0 ) {
458 Reinit();
459 }
460 else {
461 AllocBeforeWrite(nSrcLen);
2bb67b80 462 memcpy(m_pchData, pszSrcData, nSrcLen*sizeof(wxChar));
c801d85f 463 GetStringData()->nDataLength = nSrcLen;
2bb67b80 464 m_pchData[nSrcLen] = _T('\0');
c801d85f
KB
465 }
466}
467
468// assigns one string to another
469wxString& wxString::operator=(const wxString& stringSrc)
470{
097c080b
VZ
471 wxASSERT( stringSrc.GetStringData()->IsValid() );
472
c801d85f
KB
473 // don't copy string over itself
474 if ( m_pchData != stringSrc.m_pchData ) {
475 if ( stringSrc.GetStringData()->IsEmpty() ) {
476 Reinit();
477 }
478 else {
479 // adjust references
480 GetStringData()->Unlock();
481 m_pchData = stringSrc.m_pchData;
482 GetStringData()->Lock();
483 }
484 }
485
486 return *this;
487}
488
489// assigns a single character
2bb67b80 490wxString& wxString::operator=(wxChar ch)
c801d85f
KB
491{
492 AssignCopy(1, &ch);
493 return *this;
494}
495
496// assigns C string
2bb67b80 497wxString& wxString::operator=(const wxChar *psz)
c801d85f 498{
2bb67b80 499 AssignCopy(wxStrlen(psz), psz);
c801d85f
KB
500 return *this;
501}
502
2bb67b80
OK
503#if !wxUSE_UNICODE
504
c801d85f
KB
505// same as 'signed char' variant
506wxString& wxString::operator=(const unsigned char* psz)
507{
508 *this = (const char *)psz;
509 return *this;
510}
511
0f3e3e0c 512#if wxUSE_WCHAR_T
c801d85f
KB
513wxString& wxString::operator=(const wchar_t *pwz)
514{
515 wxString str(pwz);
516 *this = str;
517 return *this;
518}
0f3e3e0c 519#endif
c801d85f 520
2bb67b80
OK
521#endif
522
c801d85f
KB
523// ---------------------------------------------------------------------------
524// string concatenation
525// ---------------------------------------------------------------------------
526
c801d85f 527// add something to this string
2bb67b80 528void wxString::ConcatSelf(int nSrcLen, const wxChar *pszSrcData)
c801d85f 529{
3168a13f 530 STATISTICS_ADD(SummandLength, nSrcLen);
c801d85f 531
05488905
VZ
532 // concatenating an empty string is a NOP
533 if ( nSrcLen > 0 ) {
534 wxStringData *pData = GetStringData();
535 size_t nLen = pData->nDataLength;
536 size_t nNewLen = nLen + nSrcLen;
c801d85f 537
05488905
VZ
538 // alloc new buffer if current is too small
539 if ( pData->IsShared() ) {
540 STATISTICS_ADD(ConcatHit, 0);
3168a13f 541
05488905
VZ
542 // we have to allocate another buffer
543 wxStringData* pOldData = GetStringData();
544 AllocBuffer(nNewLen);
2bb67b80 545 memcpy(m_pchData, pOldData->data(), nLen*sizeof(wxChar));
05488905
VZ
546 pOldData->Unlock();
547 }
548 else if ( nNewLen > pData->nAllocLength ) {
549 STATISTICS_ADD(ConcatHit, 0);
3168a13f 550
05488905
VZ
551 // we have to grow the buffer
552 Alloc(nNewLen);
553 }
554 else {
555 STATISTICS_ADD(ConcatHit, 1);
3168a13f 556
05488905
VZ
557 // the buffer is already big enough
558 }
3168a13f 559
05488905
VZ
560 // should be enough space
561 wxASSERT( nNewLen <= GetStringData()->nAllocLength );
3168a13f 562
05488905 563 // fast concatenation - all is done in our buffer
2bb67b80 564 memcpy(m_pchData + nLen, pszSrcData, nSrcLen*sizeof(wxChar));
3168a13f 565
2bb67b80 566 m_pchData[nNewLen] = _T('\0'); // put terminating '\0'
05488905
VZ
567 GetStringData()->nDataLength = nNewLen; // and fix the length
568 }
569 //else: the string to append was empty
c801d85f
KB
570}
571
572/*
c801d85f
KB
573 * concatenation functions come in 5 flavours:
574 * string + string
575 * char + string and string + char
576 * C str + string and string + C str
577 */
578
579wxString operator+(const wxString& string1, const wxString& string2)
580{
097c080b
VZ
581 wxASSERT( string1.GetStringData()->IsValid() );
582 wxASSERT( string2.GetStringData()->IsValid() );
583
3168a13f
VZ
584 wxString s = string1;
585 s += string2;
586
c801d85f
KB
587 return s;
588}
589
2bb67b80 590wxString operator+(const wxString& string, wxChar ch)
c801d85f 591{
3168a13f
VZ
592 wxASSERT( string.GetStringData()->IsValid() );
593
594 wxString s = string;
595 s += ch;
097c080b 596
c801d85f
KB
597 return s;
598}
599
2bb67b80 600wxString operator+(wxChar ch, const wxString& string)
c801d85f 601{
097c080b
VZ
602 wxASSERT( string.GetStringData()->IsValid() );
603
3168a13f
VZ
604 wxString s = ch;
605 s += string;
606
c801d85f
KB
607 return s;
608}
609
2bb67b80 610wxString operator+(const wxString& string, const wxChar *psz)
c801d85f 611{
097c080b
VZ
612 wxASSERT( string.GetStringData()->IsValid() );
613
c801d85f 614 wxString s;
2bb67b80 615 s.Alloc(wxStrlen(psz) + string.Len());
3168a13f
VZ
616 s = string;
617 s += psz;
618
c801d85f
KB
619 return s;
620}
621
2bb67b80 622wxString operator+(const wxChar *psz, const wxString& string)
c801d85f 623{
097c080b
VZ
624 wxASSERT( string.GetStringData()->IsValid() );
625
c801d85f 626 wxString s;
2bb67b80 627 s.Alloc(wxStrlen(psz) + string.Len());
3168a13f
VZ
628 s = psz;
629 s += string;
630
c801d85f
KB
631 return s;
632}
633
634// ===========================================================================
635// other common string functions
636// ===========================================================================
637
638// ---------------------------------------------------------------------------
639// simple sub-string extraction
640// ---------------------------------------------------------------------------
641
642// helper function: clone the data attached to this string
643void wxString::AllocCopy(wxString& dest, int nCopyLen, int nCopyIndex) const
644{
3168a13f 645 if ( nCopyLen == 0 ) {
c801d85f
KB
646 dest.Init();
647 }
3168a13f 648 else {
c801d85f 649 dest.AllocBuffer(nCopyLen);
2bb67b80 650 memcpy(dest.m_pchData, m_pchData + nCopyIndex, nCopyLen*sizeof(wxChar));
c801d85f
KB
651 }
652}
653
654// extract string of length nCount starting at nFirst
c801d85f
KB
655wxString wxString::Mid(size_t nFirst, size_t nCount) const
656{
30d9011f
VZ
657 wxStringData *pData = GetStringData();
658 size_t nLen = pData->nDataLength;
659
566b84d2
VZ
660 // default value of nCount is wxSTRING_MAXLEN and means "till the end"
661 if ( nCount == wxSTRING_MAXLEN )
30d9011f
VZ
662 {
663 nCount = nLen - nFirst;
664 }
665
c801d85f 666 // out-of-bounds requests return sensible things
30d9011f
VZ
667 if ( nFirst + nCount > nLen )
668 {
669 nCount = nLen - nFirst;
670 }
c801d85f 671
30d9011f
VZ
672 if ( nFirst > nLen )
673 {
674 // AllocCopy() will return empty string
c801d85f 675 nCount = 0;
30d9011f 676 }
c801d85f
KB
677
678 wxString dest;
679 AllocCopy(dest, nCount, nFirst);
30d9011f 680
c801d85f
KB
681 return dest;
682}
683
684// extract nCount last (rightmost) characters
685wxString wxString::Right(size_t nCount) const
686{
687 if ( nCount > (size_t)GetStringData()->nDataLength )
688 nCount = GetStringData()->nDataLength;
689
690 wxString dest;
691 AllocCopy(dest, nCount, GetStringData()->nDataLength - nCount);
692 return dest;
693}
694
695// get all characters after the last occurence of ch
696// (returns the whole string if ch not found)
2bb67b80 697wxString wxString::AfterLast(wxChar ch) const
c801d85f
KB
698{
699 wxString str;
700 int iPos = Find(ch, TRUE);
3c67202d 701 if ( iPos == wxNOT_FOUND )
c801d85f
KB
702 str = *this;
703 else
c8cfb486 704 str = c_str() + iPos + 1;
c801d85f
KB
705
706 return str;
707}
708
709// extract nCount first (leftmost) characters
710wxString wxString::Left(size_t nCount) const
711{
712 if ( nCount > (size_t)GetStringData()->nDataLength )
713 nCount = GetStringData()->nDataLength;
714
715 wxString dest;
716 AllocCopy(dest, nCount, 0);
717 return dest;
718}
719
720// get all characters before the first occurence of ch
721// (returns the whole string if ch not found)
2bb67b80 722wxString wxString::BeforeFirst(wxChar ch) const
c801d85f
KB
723{
724 wxString str;
2bb67b80 725 for ( const wxChar *pc = m_pchData; *pc != _T('\0') && *pc != ch; pc++ )
c801d85f
KB
726 str += *pc;
727
728 return str;
729}
730
731/// get all characters before the last occurence of ch
732/// (returns empty string if ch not found)
2bb67b80 733wxString wxString::BeforeLast(wxChar ch) const
c801d85f
KB
734{
735 wxString str;
736 int iPos = Find(ch, TRUE);
3c67202d 737 if ( iPos != wxNOT_FOUND && iPos != 0 )
d1c9bbf6 738 str = wxString(c_str(), iPos);
c801d85f
KB
739
740 return str;
741}
742
743/// get all characters after the first occurence of ch
744/// (returns empty string if ch not found)
2bb67b80 745wxString wxString::AfterFirst(wxChar ch) const
c801d85f
KB
746{
747 wxString str;
748 int iPos = Find(ch);
3c67202d 749 if ( iPos != wxNOT_FOUND )
c801d85f
KB
750 str = c_str() + iPos + 1;
751
752 return str;
753}
754
755// replace first (or all) occurences of some substring with another one
2bb67b80 756size_t wxString::Replace(const wxChar *szOld, const wxChar *szNew, bool bReplaceAll)
c801d85f 757{
c86f1403 758 size_t uiCount = 0; // count of replacements made
c801d85f 759
2bb67b80 760 size_t uiOldLen = wxStrlen(szOld);
c801d85f
KB
761
762 wxString strTemp;
2bb67b80
OK
763 const wxChar *pCurrent = m_pchData;
764 const wxChar *pSubstr;
765 while ( *pCurrent != _T('\0') ) {
766 pSubstr = wxStrstr(pCurrent, szOld);
c801d85f
KB
767 if ( pSubstr == NULL ) {
768 // strTemp is unused if no replacements were made, so avoid the copy
769 if ( uiCount == 0 )
770 return 0;
771
772 strTemp += pCurrent; // copy the rest
773 break; // exit the loop
774 }
775 else {
776 // take chars before match
777 strTemp.ConcatSelf(pSubstr - pCurrent, pCurrent);
778 strTemp += szNew;
779 pCurrent = pSubstr + uiOldLen; // restart after match
780
781 uiCount++;
782
783 // stop now?
784 if ( !bReplaceAll ) {
785 strTemp += pCurrent; // copy the rest
786 break; // exit the loop
787 }
788 }
789 }
790
791 // only done if there were replacements, otherwise would have returned above
792 *this = strTemp;
793
794 return uiCount;
795}
796
797bool wxString::IsAscii() const
798{
2bb67b80 799 const wxChar *s = (const wxChar*) *this;
c801d85f
KB
800 while(*s){
801 if(!isascii(*s)) return(FALSE);
802 s++;
803 }
804 return(TRUE);
805}
dd1eaa89 806
c801d85f
KB
807bool wxString::IsWord() const
808{
2bb67b80 809 const wxChar *s = (const wxChar*) *this;
c801d85f 810 while(*s){
2bb67b80 811 if(!wxIsalpha(*s)) return(FALSE);
c801d85f
KB
812 s++;
813 }
814 return(TRUE);
815}
dd1eaa89 816
c801d85f
KB
817bool wxString::IsNumber() const
818{
2bb67b80 819 const wxChar *s = (const wxChar*) *this;
c801d85f 820 while(*s){
2bb67b80 821 if(!wxIsdigit(*s)) return(FALSE);
c801d85f
KB
822 s++;
823 }
824 return(TRUE);
825}
826
c801d85f
KB
827wxString wxString::Strip(stripType w) const
828{
829 wxString s = *this;
830 if ( w & leading ) s.Trim(FALSE);
831 if ( w & trailing ) s.Trim(TRUE);
832 return s;
833}
834
c801d85f
KB
835// ---------------------------------------------------------------------------
836// case conversion
837// ---------------------------------------------------------------------------
838
839wxString& wxString::MakeUpper()
840{
841 CopyBeforeWrite();
842
2bb67b80
OK
843 for ( wxChar *p = m_pchData; *p; p++ )
844 *p = (wxChar)wxToupper(*p);
c801d85f
KB
845
846 return *this;
847}
848
849wxString& wxString::MakeLower()
850{
851 CopyBeforeWrite();
dd1eaa89 852
2bb67b80
OK
853 for ( wxChar *p = m_pchData; *p; p++ )
854 *p = (wxChar)wxTolower(*p);
c801d85f
KB
855
856 return *this;
857}
858
859// ---------------------------------------------------------------------------
860// trimming and padding
861// ---------------------------------------------------------------------------
862
863// trims spaces (in the sense of isspace) from left or right side
864wxString& wxString::Trim(bool bFromRight)
865{
2c3b684c 866 // first check if we're going to modify the string at all
913df6f2
DW
867 if ( !IsEmpty() &&
868 (
2bb67b80
OK
869 (bFromRight && wxIsspace(GetChar(Len() - 1))) ||
870 (!bFromRight && wxIsspace(GetChar(0u)))
2c3b684c
VZ
871 )
872 )
c801d85f 873 {
2c3b684c
VZ
874 // ok, there is at least one space to trim
875 CopyBeforeWrite();
876
877 if ( bFromRight )
878 {
879 // find last non-space character
2bb67b80
OK
880 wxChar *psz = m_pchData + GetStringData()->nDataLength - 1;
881 while ( wxIsspace(*psz) && (psz >= m_pchData) )
2c3b684c
VZ
882 psz--;
883
884 // truncate at trailing space start
2bb67b80 885 *++psz = _T('\0');
2c3b684c
VZ
886 GetStringData()->nDataLength = psz - m_pchData;
887 }
888 else
889 {
890 // find first non-space character
2bb67b80
OK
891 const wxChar *psz = m_pchData;
892 while ( wxIsspace(*psz) )
2c3b684c
VZ
893 psz++;
894
895 // fix up data and length
2bb67b80
OK
896 int nDataLength = GetStringData()->nDataLength - (psz - (const wxChar*) m_pchData);
897 memmove(m_pchData, psz, (nDataLength + 1)*sizeof(wxChar));
2c3b684c
VZ
898 GetStringData()->nDataLength = nDataLength;
899 }
c801d85f
KB
900 }
901
902 return *this;
903}
904
905// adds nCount characters chPad to the string from either side
2bb67b80 906wxString& wxString::Pad(size_t nCount, wxChar chPad, bool bFromRight)
c801d85f
KB
907{
908 wxString s(chPad, nCount);
909
910 if ( bFromRight )
911 *this += s;
912 else
913 {
914 s += *this;
915 *this = s;
916 }
917
918 return *this;
919}
920
921// truncate the string
922wxString& wxString::Truncate(size_t uiLen)
923{
79a773ba
VZ
924 if ( uiLen < Len() ) {
925 CopyBeforeWrite();
926
2bb67b80 927 *(m_pchData + uiLen) = _T('\0');
79a773ba
VZ
928 GetStringData()->nDataLength = uiLen;
929 }
930 //else: nothing to do, string is already short enough
c801d85f
KB
931
932 return *this;
933}
934
935// ---------------------------------------------------------------------------
3c67202d 936// finding (return wxNOT_FOUND if not found and index otherwise)
c801d85f
KB
937// ---------------------------------------------------------------------------
938
939// find a character
2bb67b80 940int wxString::Find(wxChar ch, bool bFromEnd) const
c801d85f 941{
2bb67b80 942 const wxChar *psz = bFromEnd ? wxStrrchr(m_pchData, ch) : wxStrchr(m_pchData, ch);
c801d85f 943
2bb67b80 944 return (psz == NULL) ? wxNOT_FOUND : psz - (const wxChar*) m_pchData;
c801d85f
KB
945}
946
947// find a sub-string (like strstr)
2bb67b80 948int wxString::Find(const wxChar *pszSub) const
c801d85f 949{
2bb67b80 950 const wxChar *psz = wxStrstr(m_pchData, pszSub);
c801d85f 951
2bb67b80 952 return (psz == NULL) ? wxNOT_FOUND : psz - (const wxChar*) m_pchData;
c801d85f
KB
953}
954
7be07660
VZ
955// ---------------------------------------------------------------------------
956// stream-like operators
957// ---------------------------------------------------------------------------
958wxString& wxString::operator<<(int i)
959{
960 wxString res;
2bb67b80 961 res.Printf(_T("%d"), i);
7be07660
VZ
962
963 return (*this) << res;
964}
965
966wxString& wxString::operator<<(float f)
967{
968 wxString res;
2bb67b80 969 res.Printf(_T("%f"), f);
7be07660
VZ
970
971 return (*this) << res;
972}
973
974wxString& wxString::operator<<(double d)
975{
976 wxString res;
2bb67b80 977 res.Printf(_T("%g"), d);
7be07660
VZ
978
979 return (*this) << res;
980}
981
c801d85f 982// ---------------------------------------------------------------------------
9efd3367 983// formatted output
c801d85f 984// ---------------------------------------------------------------------------
2bb67b80 985int wxString::Printf(const wxChar *pszFormat, ...)
c801d85f
KB
986{
987 va_list argptr;
988 va_start(argptr, pszFormat);
989
990 int iLen = PrintfV(pszFormat, argptr);
991
992 va_end(argptr);
993
994 return iLen;
995}
996
2bb67b80 997int wxString::PrintfV(const wxChar* pszFormat, va_list argptr)
c801d85f 998{
7be07660 999 // static buffer to avoid dynamic memory allocation each time
11914352
KB
1000 char s_szScratch[1024]; // using static buffer causes internal compiler err
1001#if 0
2bb67b80
OK
1002#if wxUSE_THREADS
1003 // protect the static buffer
1004 static wxCriticalSection critsect;
1005 wxCriticalSectionLocker lock(critsect);
1006#endif
11914352 1007#endif
c801d85f 1008
11914352
KB
1009#if wxUSE_EXPERIMENTAL_PRINTF
1010// the new implementation
2bb67b80
OK
1011
1012 Reinit();
1013 for (size_t n = 0; pszFormat[n]; n++)
1014 if (pszFormat[n] == _T('%')) {
1015 static char s_szFlags[256] = "%";
1016 size_t flagofs = 1;
1017 bool adj_left = FALSE, in_prec = FALSE,
4e57b0d4 1018 prec_dot = FALSE, done = FALSE;
2bb67b80
OK
1019 int ilen = 0;
1020 size_t min_width = 0, max_width = wxSTRING_MAXLEN;
1021 do {
1022#define CHECK_PREC if (in_prec && !prec_dot) { s_szFlags[flagofs++] = '.'; prec_dot = TRUE; }
4e57b0d4
VZ
1023 switch (pszFormat[++n]) {
1024 case _T('\0'):
1025 done = TRUE;
1026 break;
1027 case _T('%'):
1028 *this += _T('%');
1029 done = TRUE;
1030 break;
1031 case _T('#'):
1032 case _T('0'):
1033 case _T(' '):
1034 case _T('+'):
1035 case _T('\''):
1036 CHECK_PREC
1037 s_szFlags[flagofs++] = pszFormat[n];
1038 break;
1039 case _T('-'):
1040 CHECK_PREC
1041 adj_left = TRUE;
1042 s_szFlags[flagofs++] = pszFormat[n];
1043 break;
1044 case _T('.'):
1045 CHECK_PREC
1046 in_prec = TRUE;
1047 prec_dot = FALSE;
1048 max_width = 0;
1049 // dot will be auto-added to s_szFlags if non-negative number follows
1050 break;
1051 case _T('h'):
1052 ilen = -1;
1053 CHECK_PREC
1054 s_szFlags[flagofs++] = pszFormat[n];
1055 break;
1056 case _T('l'):
1057 ilen = 1;
1058 CHECK_PREC
1059 s_szFlags[flagofs++] = pszFormat[n];
1060 break;
1061 case _T('q'):
1062 case _T('L'):
1063 ilen = 2;
1064 CHECK_PREC
1065 s_szFlags[flagofs++] = pszFormat[n];
1066 break;
1067 case _T('Z'):
1068 ilen = 3;
1069 CHECK_PREC
1070 s_szFlags[flagofs++] = pszFormat[n];
1071 break;
1072 case _T('*'):
1073 {
1074 int len = va_arg(argptr, int);
1075 if (in_prec) {
1076 if (len<0) break;
1077 CHECK_PREC
1078 max_width = len;
1079 } else {
1080 if (len<0) {
1081 adj_left = !adj_left;
1082 s_szFlags[flagofs++] = '-';
1083 len = -len;
1084 }
1085 min_width = len;
1086 }
1087 flagofs += ::sprintf(s_szFlags+flagofs,"%d",len);
1088 }
1089 break;
1090 case _T('1'): case _T('2'): case _T('3'):
1091 case _T('4'): case _T('5'): case _T('6'):
1092 case _T('7'): case _T('8'): case _T('9'):
1093 {
1094 int len = 0;
1095 CHECK_PREC
1096 while ((pszFormat[n]>=_T('0')) && (pszFormat[n]<=_T('9'))) {
1097 s_szFlags[flagofs++] = pszFormat[n];
1098 len = len*10 + (pszFormat[n] - _T('0'));
1099 n++;
1100 }
1101 if (in_prec) max_width = len;
1102 else min_width = len;
1103 n--; // the main loop pre-increments n again
1104 }
1105 break;
1106 case _T('d'):
1107 case _T('i'):
1108 case _T('o'):
1109 case _T('u'):
1110 case _T('x'):
1111 case _T('X'):
1112 CHECK_PREC
1113 s_szFlags[flagofs++] = pszFormat[n];
1114 s_szFlags[flagofs] = '\0';
1115 if (ilen == 0 ) {
1116 int val = va_arg(argptr, int);
1117 ::sprintf(s_szScratch, s_szFlags, val);
1118 }
1119 else if (ilen == -1) {
1120 short int val = va_arg(argptr, short int);
1121 ::sprintf(s_szScratch, s_szFlags, val);
1122 }
1123 else if (ilen == 1) {
1124 long int val = va_arg(argptr, long int);
1125 ::sprintf(s_szScratch, s_szFlags, val);
1126 }
1127 else if (ilen == 2) {
2bb67b80 1128#if SIZEOF_LONG_LONG
4e57b0d4
VZ
1129 long long int val = va_arg(argptr, long long int);
1130 ::sprintf(s_szScratch, s_szFlags, val);
2bb67b80 1131#else
4e57b0d4
VZ
1132 long int val = va_arg(argptr, long int);
1133 ::sprintf(s_szScratch, s_szFlags, val);
2bb67b80 1134#endif
4e57b0d4
VZ
1135 }
1136 else if (ilen == 3) {
1137 size_t val = va_arg(argptr, size_t);
1138 ::sprintf(s_szScratch, s_szFlags, val);
1139 }
1140 *this += wxString(s_szScratch);
1141 done = TRUE;
1142 break;
1143 case _T('e'):
1144 case _T('E'):
1145 case _T('f'):
1146 case _T('g'):
1147 case _T('G'):
1148 CHECK_PREC
1149 s_szFlags[flagofs++] = pszFormat[n];
1150 s_szFlags[flagofs] = '\0';
1151 if (ilen == 2) {
1152 long double val = va_arg(argptr, long double);
1153 ::sprintf(s_szScratch, s_szFlags, val);
1154 } else {
1155 double val = va_arg(argptr, double);
1156 ::sprintf(s_szScratch, s_szFlags, val);
1157 }
1158 *this += wxString(s_szScratch);
1159 done = TRUE;
1160 break;
1161 case _T('p'):
1162 {
1163 void *val = va_arg(argptr, void *);
1164 CHECK_PREC
1165 s_szFlags[flagofs++] = pszFormat[n];
1166 s_szFlags[flagofs] = '\0';
1167 ::sprintf(s_szScratch, s_szFlags, val);
1168 *this += wxString(s_szScratch);
1169 done = TRUE;
1170 }
1171 break;
1172 case _T('c'):
1173 {
1174 wxChar val = va_arg(argptr, int);
1175 // we don't need to honor padding here, do we?
1176 *this += val;
1177 done = TRUE;
1178 }
1179 break;
1180 case _T('s'):
1181 if (ilen == -1) {
1182 // wx extension: we'll let %hs mean non-Unicode strings
1183 char *val = va_arg(argptr, char *);
2bb67b80 1184#if wxUSE_UNICODE
4e57b0d4
VZ
1185 // ASCII->Unicode constructor handles max_width right
1186 wxString s(val, wxConvLibc, max_width);
2bb67b80 1187#else
4e57b0d4
VZ
1188 size_t len = wxSTRING_MAXLEN;
1189 if (val) {
1190 for (len = 0; val[len] && (len<max_width); len++);
1191 } else val = _T("(null)");
1192 wxString s(val, len);
2bb67b80 1193#endif
4e57b0d4
VZ
1194 if (s.Len() < min_width)
1195 s.Pad(min_width - s.Len(), _T(' '), adj_left);
1196 *this += s;
1197 } else {
1198 wxChar *val = va_arg(argptr, wxChar *);
1199 size_t len = wxSTRING_MAXLEN;
1200 if (val) {
1201 for (len = 0; val[len] && (len<max_width); len++);
1202 } else val = _T("(null)");
1203 wxString s(val, len);
1204 if (s.Len() < min_width)
1205 s.Pad(min_width - s.Len(), _T(' '), adj_left);
1206 *this += s;
1207 }
1208 done = TRUE;
1209 break;
1210 case _T('n'):
1211 if (ilen == 0) {
1212 int *val = va_arg(argptr, int *);
1213 *val = Len();
1214 }
1215 else if (ilen == -1) {
1216 short int *val = va_arg(argptr, short int *);
1217 *val = Len();
1218 }
1219 else if (ilen >= 1) {
1220 long int *val = va_arg(argptr, long int *);
1221 *val = Len();
1222 }
1223 done = TRUE;
1224 break;
1225 default:
1226 if (wxIsalpha(pszFormat[n]))
1227 // probably some flag not taken care of here yet
1228 s_szFlags[flagofs++] = pszFormat[n];
1229 else {
1230 // bad format
1231 *this += _T('%'); // just to pass the glibc tst-printf.c
1232 n--;
1233 done = TRUE;
1234 }
1235 break;
1236 }
2bb67b80
OK
1237#undef CHECK_PREC
1238 } while (!done);
1239 } else *this += pszFormat[n];
1240
1241#else
1242 // NB: wxVsnprintf() may return either less than the buffer size or -1 if there
89b892a2 1243 // is not enough place depending on implementation
2bb67b80 1244 int iLen = wxVsnprintf(s_szScratch, WXSIZEOF(s_szScratch), pszFormat, argptr);
7be07660 1245 char *buffer;
89b892a2 1246 if ( iLen < (int)WXSIZEOF(s_szScratch) ) {
7be07660
VZ
1247 buffer = s_szScratch;
1248 }
1249 else {
1250 int size = WXSIZEOF(s_szScratch) * 2;
1251 buffer = (char *)malloc(size);
1252 while ( buffer != NULL ) {
2bb67b80 1253 iLen = wxVsnprintf(buffer, WXSIZEOF(s_szScratch), pszFormat, argptr);
7be07660
VZ
1254 if ( iLen < size ) {
1255 // ok, there was enough space
1256 break;
1257 }
1258
1259 // still not enough, double it again
1260 buffer = (char *)realloc(buffer, size *= 2);
1261 }
1262
1263 if ( !buffer ) {
1264 // out of memory
1265 return -1;
1266 }
1267 }
1268
2bb67b80
OK
1269 wxString s(buffer);
1270 *this = s;
7be07660
VZ
1271
1272 if ( buffer != s_szScratch )
1273 free(buffer);
2bb67b80 1274#endif
c801d85f 1275
2bb67b80 1276 return Len();
c801d85f
KB
1277}
1278
097c080b
VZ
1279// ----------------------------------------------------------------------------
1280// misc other operations
1281// ----------------------------------------------------------------------------
2bb67b80 1282bool wxString::Matches(const wxChar *pszMask) const
097c080b
VZ
1283{
1284 // check char by char
2bb67b80
OK
1285 const wxChar *pszTxt;
1286 for ( pszTxt = c_str(); *pszMask != _T('\0'); pszMask++, pszTxt++ ) {
097c080b 1287 switch ( *pszMask ) {
2bb67b80
OK
1288 case _T('?'):
1289 if ( *pszTxt == _T('\0') )
097c080b
VZ
1290 return FALSE;
1291
1292 pszTxt++;
1293 pszMask++;
1294 break;
1295
2bb67b80 1296 case _T('*'):
097c080b
VZ
1297 {
1298 // ignore special chars immediately following this one
2bb67b80 1299 while ( *pszMask == _T('*') || *pszMask == _T('?') )
097c080b
VZ
1300 pszMask++;
1301
1302 // if there is nothing more, match
2bb67b80 1303 if ( *pszMask == _T('\0') )
097c080b
VZ
1304 return TRUE;
1305
1306 // are there any other metacharacters in the mask?
c86f1403 1307 size_t uiLenMask;
2bb67b80 1308 const wxChar *pEndMask = wxStrpbrk(pszMask, _T("*?"));
097c080b
VZ
1309
1310 if ( pEndMask != NULL ) {
1311 // we have to match the string between two metachars
1312 uiLenMask = pEndMask - pszMask;
1313 }
1314 else {
1315 // we have to match the remainder of the string
2bb67b80 1316 uiLenMask = wxStrlen(pszMask);
097c080b
VZ
1317 }
1318
1319 wxString strToMatch(pszMask, uiLenMask);
2bb67b80 1320 const wxChar* pMatch = wxStrstr(pszTxt, strToMatch);
097c080b
VZ
1321 if ( pMatch == NULL )
1322 return FALSE;
1323
1324 // -1 to compensate "++" in the loop
1325 pszTxt = pMatch + uiLenMask - 1;
1326 pszMask += uiLenMask - 1;
1327 }
1328 break;
1329
1330 default:
1331 if ( *pszMask != *pszTxt )
1332 return FALSE;
1333 break;
1334 }
1335 }
1336
1337 // match only if nothing left
2bb67b80 1338 return *pszTxt == _T('\0');
097c080b
VZ
1339}
1340
1fc5dd6f 1341// Count the number of chars
2bb67b80 1342int wxString::Freq(wxChar ch) const
1fc5dd6f
JS
1343{
1344 int count = 0;
1345 int len = Len();
1346 for (int i = 0; i < len; i++)
1347 {
1348 if (GetChar(i) == ch)
1349 count ++;
1350 }
1351 return count;
1352}
1353
03ab016d
JS
1354// convert to upper case, return the copy of the string
1355wxString wxString::Upper() const
1356{ wxString s(*this); return s.MakeUpper(); }
1357
1358// convert to lower case, return the copy of the string
1359wxString wxString::Lower() const { wxString s(*this); return s.MakeLower(); }
1360
2bb67b80 1361int wxString::sprintf(const wxChar *pszFormat, ...)
8870c26e
JS
1362 {
1363 va_list argptr;
1364 va_start(argptr, pszFormat);
1365 int iLen = PrintfV(pszFormat, argptr);
1366 va_end(argptr);
1367 return iLen;
1368 }
1369
c801d85f
KB
1370// ---------------------------------------------------------------------------
1371// standard C++ library string functions
1372// ---------------------------------------------------------------------------
8de2e39c 1373#ifdef wxSTD_STRING_COMPATIBILITY
c801d85f
KB
1374
1375wxString& wxString::insert(size_t nPos, const wxString& str)
1376{
097c080b 1377 wxASSERT( str.GetStringData()->IsValid() );
c801d85f
KB
1378 wxASSERT( nPos <= Len() );
1379
cb6780ff
VZ
1380 if ( !str.IsEmpty() ) {
1381 wxString strTmp;
2bb67b80
OK
1382 wxChar *pc = strTmp.GetWriteBuf(Len() + str.Len());
1383 wxStrncpy(pc, c_str(), nPos);
1384 wxStrcpy(pc + nPos, str);
1385 wxStrcpy(pc + nPos + str.Len(), c_str() + nPos);
cb6780ff
VZ
1386 strTmp.UngetWriteBuf();
1387 *this = strTmp;
1388 }
dd1eaa89
VZ
1389
1390 return *this;
c801d85f
KB
1391}
1392
1393size_t wxString::find(const wxString& str, size_t nStart) const
1394{
097c080b 1395 wxASSERT( str.GetStringData()->IsValid() );
c801d85f
KB
1396 wxASSERT( nStart <= Len() );
1397
2bb67b80 1398 const wxChar *p = wxStrstr(c_str() + nStart, str);
dd1eaa89 1399
c801d85f
KB
1400 return p == NULL ? npos : p - c_str();
1401}
1402
f0b3249b 1403// VC++ 1.5 can't cope with the default argument in the header.
3f4a0c5b 1404#if !defined(__VISUALC__) || defined(__WIN32__)
2bb67b80 1405size_t wxString::find(const wxChar* sz, size_t nStart, size_t n) const
c801d85f
KB
1406{
1407 return find(wxString(sz, n == npos ? 0 : n), nStart);
1408}
3f4a0c5b 1409#endif // VC++ 1.5
dd1eaa89 1410
62448488
JS
1411// Gives a duplicate symbol (presumably a case-insensitivity problem)
1412#if !defined(__BORLANDC__)
2bb67b80 1413size_t wxString::find(wxChar ch, size_t nStart) const
c801d85f
KB
1414{
1415 wxASSERT( nStart <= Len() );
1416
2bb67b80 1417 const wxChar *p = wxStrchr(c_str() + nStart, ch);
dd1eaa89 1418
c801d85f
KB
1419 return p == NULL ? npos : p - c_str();
1420}
62448488 1421#endif
c801d85f
KB
1422
1423size_t wxString::rfind(const wxString& str, size_t nStart) const
1424{
097c080b 1425 wxASSERT( str.GetStringData()->IsValid() );
c801d85f
KB
1426 wxASSERT( nStart <= Len() );
1427
969d318c 1428 // TODO could be made much quicker than that
2bb67b80 1429 const wxChar *p = c_str() + (nStart == npos ? Len() : nStart);
c801d85f 1430 while ( p >= c_str() + str.Len() ) {
2bb67b80 1431 if ( wxStrncmp(p - str.Len(), str, str.Len()) == 0 )
c801d85f
KB
1432 return p - str.Len() - c_str();
1433 p--;
1434 }
dd1eaa89 1435
c801d85f
KB
1436 return npos;
1437}
dd1eaa89 1438
f0b3249b 1439// VC++ 1.5 can't cope with the default argument in the header.
3f4a0c5b 1440#if !defined(__VISUALC__) || defined(__WIN32__)
2bb67b80 1441size_t wxString::rfind(const wxChar* sz, size_t nStart, size_t n) const
c801d85f 1442{
969d318c 1443 return rfind(wxString(sz, n == npos ? 0 : n), nStart);
c801d85f
KB
1444}
1445
2bb67b80 1446size_t wxString::rfind(wxChar ch, size_t nStart) const
c801d85f 1447{
969d318c
VZ
1448 if ( nStart == npos )
1449 {
1450 nStart = Len();
1451 }
1452 else
1453 {
1454 wxASSERT( nStart <= Len() );
1455 }
c801d85f 1456
969d318c 1457 const wxChar *p = wxStrrchr(c_str(), ch);
dd1eaa89 1458
969d318c
VZ
1459 if ( p == NULL )
1460 return npos;
1461
1462 size_t result = p - c_str();
1463 return ( result > nStart ) ? npos : result;
c801d85f 1464}
3f4a0c5b 1465#endif // VC++ 1.5
c801d85f 1466
969d318c
VZ
1467size_t wxString::find_first_of(const wxChar* sz, size_t nStart) const
1468{
25dd56b1
BM
1469 const wxChar *start = c_str() + nStart;
1470 const wxChar *firstOf = wxStrpbrk(start, sz);
969d318c
VZ
1471 if ( firstOf )
1472 return firstOf - start;
1473 else
1474 return npos;
1475}
1476
1477size_t wxString::find_last_of(const wxChar* sz, size_t nStart) const
1478{
1479 if ( nStart == npos )
1480 {
1481 nStart = Len();
1482 }
1483 else
1484 {
1485 wxASSERT( nStart <= Len() );
1486 }
1487
20272888 1488 for ( const wxChar *p = c_str() + length() - 1; p >= c_str(); p-- )
969d318c
VZ
1489 {
1490 if ( wxStrchr(sz, *p) )
1491 return p - c_str();
1492 }
1493
1494 return npos;
1495}
1496
1497size_t wxString::find_first_not_of(const wxChar* sz, size_t nStart) const
1498{
1499 if ( nStart == npos )
1500 {
1501 nStart = Len();
1502 }
1503 else
1504 {
1505 wxASSERT( nStart <= Len() );
1506 }
1507
20272888 1508 size_t nAccept = wxStrspn(c_str() + nStart, sz);
969d318c
VZ
1509 if ( nAccept >= length() - nStart )
1510 return npos;
1511 else
1512 return nAccept;
1513}
1514
1515size_t wxString::find_first_not_of(wxChar ch, size_t nStart) const
1516{
1517 wxASSERT( nStart <= Len() );
1518
20272888 1519 for ( const wxChar *p = c_str() + nStart; *p; p++ )
969d318c
VZ
1520 {
1521 if ( *p != ch )
1522 return p - c_str();
1523 }
1524
1525 return npos;
1526}
1527
1528size_t wxString::find_last_not_of(const wxChar* sz, size_t nStart) const
1529{
1530 if ( nStart == npos )
1531 {
1532 nStart = Len();
1533 }
1534 else
1535 {
1536 wxASSERT( nStart <= Len() );
1537 }
1538
20272888 1539 for ( const wxChar *p = c_str() + nStart - 1; p >= c_str(); p-- )
969d318c
VZ
1540 {
1541 if ( !wxStrchr(sz, *p) )
1542 return p - c_str();
1543 }
1544
1545 return npos;
1546}
1547
1548size_t wxString::find_last_not_of(wxChar ch, size_t nStart) const
1549{
1550 if ( nStart == npos )
1551 {
1552 nStart = Len();
1553 }
1554 else
1555 {
1556 wxASSERT( nStart <= Len() );
1557 }
1558
20272888 1559 for ( const wxChar *p = c_str() + nStart - 1; p >= c_str(); p-- )
969d318c
VZ
1560 {
1561 if ( *p != ch )
1562 return p - c_str();
1563 }
1564
1565 return npos;
1566}
1567
c801d85f
KB
1568wxString wxString::substr(size_t nStart, size_t nLen) const
1569{
1570 // npos means 'take all'
1571 if ( nLen == npos )
1572 nLen = 0;
1573
1574 wxASSERT( nStart + nLen <= Len() );
1575
1576 return wxString(c_str() + nStart, nLen == npos ? 0 : nLen);
1577}
1578
1579wxString& wxString::erase(size_t nStart, size_t nLen)
1580{
1581 wxString strTmp(c_str(), nStart);
1582 if ( nLen != npos ) {
1583 wxASSERT( nStart + nLen <= Len() );
1584
1585 strTmp.append(c_str() + nStart + nLen);
1586 }
1587
1588 *this = strTmp;
1589 return *this;
1590}
1591
2bb67b80 1592wxString& wxString::replace(size_t nStart, size_t nLen, const wxChar *sz)
c801d85f 1593{
2bb67b80 1594 wxASSERT( nStart + nLen <= wxStrlen(sz) );
c801d85f
KB
1595
1596 wxString strTmp;
1597 if ( nStart != 0 )
1598 strTmp.append(c_str(), nStart);
1599 strTmp += sz;
1600 strTmp.append(c_str() + nStart + nLen);
dd1eaa89 1601
c801d85f
KB
1602 *this = strTmp;
1603 return *this;
1604}
1605
2bb67b80 1606wxString& wxString::replace(size_t nStart, size_t nLen, size_t nCount, wxChar ch)
c801d85f
KB
1607{
1608 return replace(nStart, nLen, wxString(ch, nCount));
1609}
1610
dd1eaa89 1611wxString& wxString::replace(size_t nStart, size_t nLen,
097c080b 1612 const wxString& str, size_t nStart2, size_t nLen2)
c801d85f
KB
1613{
1614 return replace(nStart, nLen, str.substr(nStart2, nLen2));
1615}
1616
dd1eaa89 1617wxString& wxString::replace(size_t nStart, size_t nLen,
2bb67b80 1618 const wxChar* sz, size_t nCount)
c801d85f
KB
1619{
1620 return replace(nStart, nLen, wxString(sz, nCount));
1621}
1622
1623#endif //std::string compatibility
1624
1625// ============================================================================
1626// ArrayString
1627// ============================================================================
1628
1629// size increment = max(50% of current size, ARRAY_MAXSIZE_INCREMENT)
1630#define ARRAY_MAXSIZE_INCREMENT 4096
1631#ifndef ARRAY_DEFAULT_INITIAL_SIZE // also defined in dynarray.h
1632 #define ARRAY_DEFAULT_INITIAL_SIZE (16)
1633#endif
1634
1635#define STRING(p) ((wxString *)(&(p)))
1636
1637// ctor
1638wxArrayString::wxArrayString()
1639{
1640 m_nSize =
1641 m_nCount = 0;
2bb67b80 1642 m_pItems = (wxChar **) NULL;
c801d85f
KB
1643}
1644
1645// copy ctor
1646wxArrayString::wxArrayString(const wxArrayString& src)
1647{
3bbb630a
VZ
1648 m_nSize =
1649 m_nCount = 0;
2bb67b80 1650 m_pItems = (wxChar **) NULL;
c801d85f 1651
4d14b524 1652 *this = src;
c801d85f
KB
1653}
1654
4d14b524 1655// assignment operator
c801d85f
KB
1656wxArrayString& wxArrayString::operator=(const wxArrayString& src)
1657{
d93f63db
VZ
1658 if ( m_nSize > 0 )
1659 Clear();
c801d85f 1660
4d14b524
VZ
1661 if ( src.m_nCount > ARRAY_DEFAULT_INITIAL_SIZE )
1662 Alloc(src.m_nCount);
c801d85f 1663
4d14b524
VZ
1664 // we can't just copy the pointers here because otherwise we would share
1665 // the strings with another array
c86f1403 1666 for ( size_t n = 0; n < src.m_nCount; n++ )
4d14b524 1667 Add(src[n]);
c801d85f 1668
3bbb630a 1669 if ( m_nCount != 0 )
2bb67b80 1670 memcpy(m_pItems, src.m_pItems, m_nCount*sizeof(wxChar *));
3bbb630a 1671
c801d85f
KB
1672 return *this;
1673}
1674
1675// grow the array
1676void wxArrayString::Grow()
1677{
1678 // only do it if no more place
1679 if( m_nCount == m_nSize ) {
1680 if( m_nSize == 0 ) {
1681 // was empty, alloc some memory
1682 m_nSize = ARRAY_DEFAULT_INITIAL_SIZE;
2bb67b80 1683 m_pItems = new wxChar *[m_nSize];
c801d85f
KB
1684 }
1685 else {
3bbb630a
VZ
1686 // otherwise when it's called for the first time, nIncrement would be 0
1687 // and the array would never be expanded
913df6f2
DW
1688#if defined(__VISAGECPP__)
1689 int array_size = ARRAY_DEFAULT_INITIAL_SIZE;
1690 wxASSERT( array_size != 0 );
1691#else
3bbb630a 1692 wxASSERT( ARRAY_DEFAULT_INITIAL_SIZE != 0 );
913df6f2 1693#endif
3bbb630a 1694
c801d85f 1695 // add 50% but not too much
3bbb630a 1696 size_t nIncrement = m_nSize < ARRAY_DEFAULT_INITIAL_SIZE
4d14b524 1697 ? ARRAY_DEFAULT_INITIAL_SIZE : m_nSize >> 1;
c801d85f
KB
1698 if ( nIncrement > ARRAY_MAXSIZE_INCREMENT )
1699 nIncrement = ARRAY_MAXSIZE_INCREMENT;
1700 m_nSize += nIncrement;
2bb67b80 1701 wxChar **pNew = new wxChar *[m_nSize];
c801d85f
KB
1702
1703 // copy data to new location
2bb67b80 1704 memcpy(pNew, m_pItems, m_nCount*sizeof(wxChar *));
c801d85f
KB
1705
1706 // delete old memory (but do not release the strings!)
a3622daa 1707 wxDELETEA(m_pItems);
c801d85f
KB
1708
1709 m_pItems = pNew;
1710 }
1711 }
1712}
1713
1714void wxArrayString::Free()
1715{
1716 for ( size_t n = 0; n < m_nCount; n++ ) {
1717 STRING(m_pItems[n])->GetStringData()->Unlock();
1718 }
1719}
1720
1721// deletes all the strings from the list
1722void wxArrayString::Empty()
1723{
1724 Free();
1725
1726 m_nCount = 0;
1727}
1728
1729// as Empty, but also frees memory
1730void wxArrayString::Clear()
1731{
1732 Free();
1733
dd1eaa89 1734 m_nSize =
c801d85f
KB
1735 m_nCount = 0;
1736
a3622daa 1737 wxDELETEA(m_pItems);
c801d85f
KB
1738}
1739
1740// dtor
1741wxArrayString::~wxArrayString()
1742{
1743 Free();
1744
a3622daa 1745 wxDELETEA(m_pItems);
c801d85f
KB
1746}
1747
1748// pre-allocates memory (frees the previous data!)
1749void wxArrayString::Alloc(size_t nSize)
1750{
1751 wxASSERT( nSize > 0 );
1752
1753 // only if old buffer was not big enough
1754 if ( nSize > m_nSize ) {
1755 Free();
a3622daa 1756 wxDELETEA(m_pItems);
2bb67b80 1757 m_pItems = new wxChar *[nSize];
c801d85f
KB
1758 m_nSize = nSize;
1759 }
1760
1761 m_nCount = 0;
1762}
1763
d4ffe273
OK
1764// minimizes the memory usage by freeing unused memory
1765void wxArrayString::Shrink()
1766{
1767 // only do it if we have some memory to free
1768 if( m_nCount < m_nSize ) {
1769 // allocates exactly as much memory as we need
913df6f2 1770 wxChar **pNew = new wxChar *[m_nCount];
d4ffe273
OK
1771
1772 // copy data to new location
1773 memcpy(pNew, m_pItems, m_nCount*sizeof(wxChar *));
1774 delete [] m_pItems;
1775 m_pItems = pNew;
1776 }
1777}
1778
c801d85f 1779// searches the array for an item (forward or backwards)
2bb67b80 1780int wxArrayString::Index(const wxChar *sz, bool bCase, bool bFromEnd) const
c801d85f
KB
1781{
1782 if ( bFromEnd ) {
1783 if ( m_nCount > 0 ) {
c86f1403 1784 size_t ui = m_nCount;
c801d85f
KB
1785 do {
1786 if ( STRING(m_pItems[--ui])->IsSameAs(sz, bCase) )
1787 return ui;
1788 }
1789 while ( ui != 0 );
1790 }
1791 }
1792 else {
c86f1403 1793 for( size_t ui = 0; ui < m_nCount; ui++ ) {
c801d85f
KB
1794 if( STRING(m_pItems[ui])->IsSameAs(sz, bCase) )
1795 return ui;
1796 }
1797 }
1798
3c67202d 1799 return wxNOT_FOUND;
c801d85f
KB
1800}
1801
1802// add item at the end
097c080b 1803void wxArrayString::Add(const wxString& str)
c801d85f 1804{
097c080b
VZ
1805 wxASSERT( str.GetStringData()->IsValid() );
1806
c801d85f
KB
1807 Grow();
1808
1809 // the string data must not be deleted!
097c080b 1810 str.GetStringData()->Lock();
2bb67b80 1811 m_pItems[m_nCount++] = (wxChar *)str.c_str();
c801d85f
KB
1812}
1813
1814// add item at the given position
097c080b 1815void wxArrayString::Insert(const wxString& str, size_t nIndex)
c801d85f 1816{
097c080b
VZ
1817 wxASSERT( str.GetStringData()->IsValid() );
1818
cf2f341a 1819 wxCHECK_RET( nIndex <= m_nCount, _("bad index in wxArrayString::Insert") );
c801d85f
KB
1820
1821 Grow();
1822
dd1eaa89 1823 memmove(&m_pItems[nIndex + 1], &m_pItems[nIndex],
2bb67b80 1824 (m_nCount - nIndex)*sizeof(wxChar *));
c801d85f 1825
097c080b 1826 str.GetStringData()->Lock();
2bb67b80 1827 m_pItems[nIndex] = (wxChar *)str.c_str();
c801d85f
KB
1828
1829 m_nCount++;
1830}
1831
1832// removes item from array (by index)
1833void wxArrayString::Remove(size_t nIndex)
1834{
1a5a8367 1835 wxCHECK_RET( nIndex <= m_nCount, _("bad index in wxArrayString::Remove") );
c801d85f
KB
1836
1837 // release our lock
1838 Item(nIndex).GetStringData()->Unlock();
1839
dd1eaa89 1840 memmove(&m_pItems[nIndex], &m_pItems[nIndex + 1],
2bb67b80 1841 (m_nCount - nIndex - 1)*sizeof(wxChar *));
c801d85f
KB
1842 m_nCount--;
1843}
1844
1845// removes item from array (by value)
2bb67b80 1846void wxArrayString::Remove(const wxChar *sz)
c801d85f
KB
1847{
1848 int iIndex = Index(sz);
1849
3c67202d 1850 wxCHECK_RET( iIndex != wxNOT_FOUND,
1a5a8367 1851 _("removing inexistent element in wxArrayString::Remove") );
c801d85f 1852
c86f1403 1853 Remove(iIndex);
c801d85f
KB
1854}
1855
30b21f9a
VZ
1856// ----------------------------------------------------------------------------
1857// sorting
1858// ----------------------------------------------------------------------------
1859
1860// we can only sort one array at a time with the quick-sort based
1861// implementation
1862#if wxUSE_THREADS
30b21f9a
VZ
1863 // need a critical section to protect access to gs_compareFunction and
1864 // gs_sortAscending variables
26128999 1865 static wxCriticalSection *gs_critsectStringSort = NULL;
30b21f9a
VZ
1866
1867 // call this before the value of the global sort vars is changed/after
1868 // you're finished with them
26128999
VZ
1869 #define START_SORT() wxASSERT( !gs_critsectStringSort ); \
1870 gs_critsectStringSort = new wxCriticalSection; \
1871 gs_critsectStringSort->Enter()
1872 #define END_SORT() gs_critsectStringSort->Leave(); \
1873 delete gs_critsectStringSort; \
1874 gs_critsectStringSort = NULL
30b21f9a
VZ
1875#else // !threads
1876 #define START_SORT()
1877 #define END_SORT()
1878#endif // wxUSE_THREADS
1879
1880// function to use for string comparaison
1881static wxArrayString::CompareFunction gs_compareFunction = NULL;
1882
1883// if we don't use the compare function, this flag tells us if we sort the
1884// array in ascending or descending order
1885static bool gs_sortAscending = TRUE;
1886
1887// function which is called by quick sort
913df6f2 1888static int LINKAGEMODE wxStringCompareFunction(const void *first, const void *second)
30b21f9a
VZ
1889{
1890 wxString *strFirst = (wxString *)first;
1891 wxString *strSecond = (wxString *)second;
1892
64716cd7 1893 if ( gs_compareFunction ) {
30b21f9a 1894 return gs_compareFunction(*strFirst, *strSecond);
64716cd7 1895 }
30b21f9a 1896 else {
2bb67b80
OK
1897 // maybe we should use wxStrcoll
1898 int result = wxStrcmp(strFirst->c_str(), strSecond->c_str());
30b21f9a
VZ
1899
1900 return gs_sortAscending ? result : -result;
1901 }
1902}
1903
c801d85f 1904// sort array elements using passed comparaison function
30b21f9a
VZ
1905void wxArrayString::Sort(CompareFunction compareFunction)
1906{
1907 START_SORT();
1908
1909 wxASSERT( !gs_compareFunction ); // must have been reset to NULL
1910 gs_compareFunction = compareFunction;
1911
1912 DoSort();
1913
1914 END_SORT();
1915}
1916
1917void wxArrayString::Sort(bool reverseOrder)
1918{
1919 START_SORT();
1920
1921 wxASSERT( !gs_compareFunction ); // must have been reset to NULL
1922 gs_sortAscending = !reverseOrder;
1923
1924 DoSort();
1925
1926 END_SORT();
1927}
c801d85f 1928
30b21f9a 1929void wxArrayString::DoSort()
c801d85f 1930{
30b21f9a
VZ
1931 // just sort the pointers using qsort() - of course it only works because
1932 // wxString() *is* a pointer to its data
2bb67b80 1933 qsort(m_pItems, m_nCount, sizeof(wxChar *), wxStringCompareFunction);
c801d85f 1934}
2bb67b80
OK
1935
1936// ============================================================================
1937// MBConv
1938// ============================================================================
1939
c50c1ddc
OK
1940WXDLLEXPORT_DATA(wxMBConv *) wxConvCurrent = &wxConvLibc;
1941#if !wxUSE_WCHAR_T
1942WXDLLEXPORT_DATA(wxMBConv) wxConvLibc, wxConvFile;
1943#endif
1944
eea4f86a 1945#if wxUSE_WCHAR_T
3e473156 1946
2bb67b80
OK
1947// ----------------------------------------------------------------------------
1948// standard libc conversion
1949// ----------------------------------------------------------------------------
1950
c50c1ddc 1951WXDLLEXPORT_DATA(wxMBConv) wxConvLibc;
ba555d51 1952
6b769f3d 1953size_t wxMBConv::MB2WC(wchar_t *buf, const char *psz, size_t n) const
2bb67b80
OK
1954{
1955 return wxMB2WC(buf, psz, n);
1956}
1957
6b769f3d 1958size_t wxMBConv::WC2MB(char *buf, const wchar_t *psz, size_t n) const
2bb67b80
OK
1959{
1960 return wxWC2MB(buf, psz, n);
1961}
1962
1963// ----------------------------------------------------------------------------
ba555d51 1964// standard file conversion
2bb67b80
OK
1965// ----------------------------------------------------------------------------
1966
5f709e67 1967WXDLLEXPORT_DATA(wxMBConvFile) wxConvFile;
ba555d51
OK
1968
1969// just use the libc conversion for now
5f709e67 1970size_t wxMBConvFile::MB2WC(wchar_t *buf, const char *psz, size_t n) const
ba555d51
OK
1971{
1972 return wxMB2WC(buf, psz, n);
1973}
1974
5f709e67 1975size_t wxMBConvFile::WC2MB(char *buf, const wchar_t *psz, size_t n) const
2bb67b80 1976{
ba555d51
OK
1977 return wxWC2MB(buf, psz, n);
1978}
2bb67b80 1979
ba555d51
OK
1980// ----------------------------------------------------------------------------
1981// standard gdk conversion
1982// ----------------------------------------------------------------------------
1983
e5ea3f7a 1984#ifdef __WXGTK12__
5f709e67 1985WXDLLEXPORT_DATA(wxMBConvGdk) wxConvGdk;
ba555d51
OK
1986
1987#include <gdk/gdk.h>
1988
5f709e67 1989size_t wxMBConvGdk::MB2WC(wchar_t *buf, const char *psz, size_t n) const
ba555d51
OK
1990{
1991 if (buf) {
1992 return gdk_mbstowcs((GdkWChar *)buf, psz, n);
1993 } else {
1994 GdkWChar *nbuf = new GdkWChar[n=strlen(psz)];
1995 size_t len = gdk_mbstowcs(nbuf, psz, n);
1996 delete [] nbuf;
1997 return len;
1998 }
1999}
2000
5f709e67 2001size_t wxMBConvGdk::WC2MB(char *buf, const wchar_t *psz, size_t n) const
ba555d51
OK
2002{
2003 char *mbstr = gdk_wcstombs((GdkWChar *)psz);
2004 size_t len = mbstr ? strlen(mbstr) : 0;
2005 if (buf) {
2006 if (len > n) len = n;
2007 memcpy(buf, psz, len);
2008 if (len < n) buf[len] = 0;
2009 }
2010 return len;
2011}
7fd43a44 2012#endif // GTK > 1.0
ba555d51
OK
2013
2014// ----------------------------------------------------------------------------
2015// UTF-7
2016// ----------------------------------------------------------------------------
2017
5f709e67 2018WXDLLEXPORT_DATA(wxMBConvUTF7) wxConvUTF7;
2bb67b80 2019
eea4f86a
OK
2020#if 0
2021static char utf7_setD[]="ABCDEFGHIJKLMNOPQRSTUVWXYZ"
2022 "abcdefghijklmnopqrstuvwxyz"
2023 "0123456789'(),-./:?";
2024static char utf7_setO[]="!\"#$%&*;<=>@[]^_`{|}";
2025static char utf7_setB[]="ABCDEFGHIJKLMNOPQRSTUVWXYZ"
2026 "abcdefghijklmnopqrstuvwxyz"
2027 "0123456789+/";
2028#endif
2029
2bb67b80 2030// TODO: write actual implementations of UTF-7 here
74e3313b
VZ
2031size_t wxMBConvUTF7::MB2WC(wchar_t * WXUNUSED(buf),
2032 const char * WXUNUSED(psz),
2033 size_t WXUNUSED(n)) const
2bb67b80
OK
2034{
2035 return 0;
2036}
2037
74e3313b
VZ
2038size_t wxMBConvUTF7::WC2MB(char * WXUNUSED(buf),
2039 const wchar_t * WXUNUSED(psz),
2040 size_t WXUNUSED(n)) const
2bb67b80
OK
2041{
2042 return 0;
2043}
2044
2045// ----------------------------------------------------------------------------
2046// UTF-8
2047// ----------------------------------------------------------------------------
2048
5f709e67 2049WXDLLEXPORT_DATA(wxMBConvUTF8) wxConvUTF8;
2bb67b80 2050
eea4f86a
OK
2051static unsigned long utf8_max[]={0x7f,0x7ff,0xffff,0x1fffff,0x3ffffff,0x7fffffff,0xffffffff};
2052
5f709e67 2053size_t wxMBConvUTF8::MB2WC(wchar_t *buf, const char *psz, size_t n) const
2bb67b80 2054{
eea4f86a
OK
2055 size_t len = 0;
2056
2057 while (*psz && ((!buf) || (len<n))) {
2058 unsigned char cc=*psz++, fc=cc;
2059 unsigned cnt;
2060 for (cnt=0; fc&0x80; cnt++) fc<<=1;
2061 if (!cnt) {
2062 // plain ASCII char
2063 if (buf) *buf++=cc;
2064 len++;
2065 } else {
2066 cnt--;
2067 if (!cnt) {
4e57b0d4
VZ
2068 // invalid UTF-8 sequence
2069 return (size_t)-1;
eea4f86a 2070 } else {
4e57b0d4
VZ
2071 unsigned ocnt=cnt-1;
2072 unsigned long res=cc&(0x3f>>cnt);
2073 while (cnt--) {
2074 cc = *psz++;
2075 if ((cc&0xC0)!=0x80) {
2076 // invalid UTF-8 sequence
2077 return (size_t)-1;
2078 }
2079 res=(res<<6)|(cc&0x3f);
2080 }
2081 if (res<=utf8_max[ocnt]) {
2082 // illegal UTF-8 encoding
2083 return (size_t)-1;
2084 }
2085 if (buf) *buf++=res;
2086 len++;
eea4f86a
OK
2087 }
2088 }
2089 }
2090 if (buf && (len<n)) *buf = 0;
2091 return len;
2bb67b80
OK
2092}
2093
5f709e67 2094size_t wxMBConvUTF8::WC2MB(char *buf, const wchar_t *psz, size_t n) const
2bb67b80 2095{
eea4f86a
OK
2096 size_t len = 0;
2097
2098 while (*psz && ((!buf) || (len<n))) {
2099 unsigned long cc=(*psz++)&0x7fffffff;
2100 unsigned cnt;
2101 for (cnt=0; cc>utf8_max[cnt]; cnt++);
2102 if (!cnt) {
2103 // plain ASCII char
2104 if (buf) *buf++=cc;
2105 len++;
2106 } else {
2107 len+=cnt+1;
2108 if (buf) {
4e57b0d4
VZ
2109 *buf++=(-128>>cnt)|((cc>>(cnt*6))&(0x3f>>cnt));
2110 while (cnt--)
2111 *buf++=0x80|((cc>>(cnt*6))&0x3f);
eea4f86a
OK
2112 }
2113 }
2114 }
2115 if (buf && (len<n)) *buf = 0;
2116 return len;
2bb67b80
OK
2117}
2118
2119// ----------------------------------------------------------------------------
2120// specified character set
2121// ----------------------------------------------------------------------------
2122
3e473156
OK
2123class wxCharacterSet
2124{
2125public:
2126 wxArrayString names;
2127 wchar_t *data;
2128};
2129
2130#ifndef WX_PRECOMP
2131 #include "wx/dynarray.h"
2132 #include "wx/filefn.h"
2133 #include "wx/textfile.h"
2134 #include "wx/tokenzr.h"
2135 #include "wx/utils.h"
2136#endif
2137
2138WX_DECLARE_OBJARRAY(wxCharacterSet, wxCSArray);
2139#include "wx/arrimpl.cpp"
2140WX_DEFINE_OBJARRAY(wxCSArray);
2141
2142static wxCSArray wxCharsets;
2143
2144static void wxLoadCharacterSets(void)
2145{
2146 static bool already_loaded = FALSE;
2147
a28d23bb
OK
2148 if (already_loaded) return;
2149
eea4f86a 2150 already_loaded = TRUE;
88ac883a 2151#if defined(__UNIX__) && wxUSE_TEXTFILE
3e473156 2152 // search through files in /usr/share/i18n/charmaps
a28d23bb 2153 wxString fname;
a28d23bb 2154 for (fname = ::wxFindFirstFile(_T("/usr/share/i18n/charmaps/*"));
3e473156
OK
2155 !fname.IsEmpty();
2156 fname = ::wxFindNextFile()) {
2157 wxTextFile cmap(fname);
2158 if (cmap.Open()) {
2159 wxCharacterSet *cset = new wxCharacterSet;
2160 wxString comchar,escchar;
2161 bool in_charset = FALSE;
2162
c71c744a 2163 // wxFprintf(stderr,_T("Loaded: %s\n"),fname.c_str());
3e473156 2164
a28d23bb
OK
2165 wxString line;
2166 for (line = cmap.GetFirstLine();
4e57b0d4
VZ
2167 !cmap.Eof();
2168 line = cmap.GetNextLine()) {
2169 // wxFprintf(stderr,_T("line contents: %s\n"),line.c_str());
2170 wxStringTokenizer token(line);
2171 wxString cmd = token.GetNextToken();
2172 if (cmd == comchar) {
2173 if (token.GetNextToken() == _T("alias"))
2174 cset->names.Add(token.GetNextToken());
2175 }
2176 else if (cmd == _T("<code_set_name>"))
2177 cset->names.Add(token.GetNextToken());
2178 else if (cmd == _T("<comment_char>"))
2179 comchar = token.GetNextToken();
2180 else if (cmd == _T("<escape_char>"))
2181 escchar = token.GetNextToken();
2182 else if (cmd == _T("<mb_cur_min>")) {
2183 delete cset;
2184 cset = (wxCharacterSet *) NULL;
2185 break; // we don't support multibyte charsets ourselves (yet)
2186 }
2187 else if (cmd == _T("CHARMAP")) {
2188 cset->data = (wchar_t *)calloc(256, sizeof(wchar_t));
2189 in_charset = TRUE;
2190 }
2191 else if (cmd == _T("END")) {
2192 if (token.GetNextToken() == _T("CHARMAP"))
2193 in_charset = FALSE;
2194 }
2195 else if (in_charset) {
2196 // format: <NUL> /x00 <U0000> NULL (NUL)
2197 // <A> /x41 <U0041> LATIN CAPITAL LETTER A
2198 wxString hex = token.GetNextToken();
2199 // skip whitespace (why doesn't wxStringTokenizer do this?)
2200 while (wxIsEmpty(hex) && token.HasMoreTokens()) hex = token.GetNextToken();
2201 wxString uni = token.GetNextToken();
2202 // skip whitespace again
2203 while (wxIsEmpty(uni) && token.HasMoreTokens()) uni = token.GetNextToken();
2204 if ((hex.Len() > 2) && (wxString(hex.GetChar(0)) == escchar) && (hex.GetChar(1) == _T('x')) &&
2205 (uni.Left(2) == _T("<U"))) {
2206 hex.MakeUpper(); uni.MakeUpper();
2207 int pos = ::wxHexToDec(hex.Mid(2,2));
2208 if (pos>=0) {
2209 unsigned long uni1 = ::wxHexToDec(uni.Mid(2,2));
2210 unsigned long uni2 = ::wxHexToDec(uni.Mid(4,2));
2211 cset->data[pos] = (uni1 << 16) | uni2;
2212 // wxFprintf(stderr,_T("char %02x mapped to %04x (%c)\n"),pos,cset->data[pos],cset->data[pos]);
2213 }
2214 }
2215 }
3e473156 2216 }
c71c744a 2217 if (cset) {
4e57b0d4
VZ
2218 cset->names.Shrink();
2219 wxCharsets.Add(cset);
c71c744a 2220 }
3e473156
OK
2221 }
2222 }
2223#endif
2224 wxCharsets.Shrink();
3e473156
OK
2225}
2226
a28d23bb 2227static wxCharacterSet *wxFindCharacterSet(const wxChar *charset)
3e473156 2228{
6cf7c00f 2229 if (!charset) return (wxCharacterSet *)NULL;
a28d23bb 2230 wxLoadCharacterSets();
3e473156
OK
2231 for (size_t n=0; n<wxCharsets.GetCount(); n++)
2232 if (wxCharsets[n].names.Index(charset) != wxNOT_FOUND)
2233 return &(wxCharsets[n]);
2234 return (wxCharacterSet *)NULL;
2235}
2236
5f709e67 2237WXDLLEXPORT_DATA(wxCSConv) wxConvLocal((const wxChar *)NULL);
3e473156 2238
2bb67b80
OK
2239wxCSConv::wxCSConv(const wxChar *charset)
2240{
6cf7c00f 2241 m_name = (wxChar *) NULL;
a28d23bb 2242 m_cset = (wxCharacterSet *) NULL;
6cf7c00f
OK
2243 m_deferred = TRUE;
2244 SetName(charset);
2245}
2246
2247wxCSConv::~wxCSConv()
2248{
2249 if (m_name) free(m_name);
2250}
2251
2252void wxCSConv::SetName(const wxChar *charset)
2253{
a28d23bb 2254 if (charset) {
3e473156 2255#ifdef __UNIX__
a28d23bb
OK
2256 // first, convert the character set name to standard form
2257 wxString codeset;
2258 if (wxString(charset,3).CmpNoCase(_T("ISO")) == 0) {
2259 // make sure it's represented in the standard form: ISO_8859-1
2260 codeset = _T("ISO_");
2261 charset += 3;
2262 if ((*charset == _T('-')) || (*charset == _T('_'))) charset++;
2263 if (wxStrlen(charset)>4) {
4e57b0d4
VZ
2264 if (wxString(charset,4) == _T("8859")) {
2265 codeset << _T("8859-");
2266 if (*charset == _T('-')) charset++;
2267 }
3e473156
OK
2268 }
2269 }
a28d23bb
OK
2270 codeset << charset;
2271 codeset.MakeUpper();
2272 m_name = wxStrdup(codeset.c_str());
2273 m_deferred = TRUE;
3e473156 2274#endif
a28d23bb
OK
2275 }
2276}
2277
a28d23bb 2278void wxCSConv::LoadNow()
ba555d51 2279{
a28d23bb
OK
2280// wxPrintf(_T("Conversion request\n"));
2281 if (m_deferred) {
6cf7c00f
OK
2282 if (!m_name) {
2283#ifdef __UNIX__
2284 wxChar *lang = wxGetenv(_T("LANG"));
2285 wxChar *dot = lang ? wxStrchr(lang, _T('.')) : (wxChar *)NULL;
2286 if (dot) SetName(dot+1);
2287#endif
2288 }
a28d23bb
OK
2289 m_cset = wxFindCharacterSet(m_name);
2290 m_deferred = FALSE;
2291 }
ba555d51
OK
2292}
2293
6b769f3d 2294size_t wxCSConv::MB2WC(wchar_t *buf, const char *psz, size_t n) const
2bb67b80 2295{
a28d23bb 2296 ((wxCSConv *)this)->LoadNow(); // discard constness
3e473156 2297 if (buf) {
a28d23bb 2298 if (m_cset) {
38444daf 2299 for (size_t c=0; c<n; c++)
4e57b0d4 2300 buf[c] = m_cset->data[(unsigned char)(psz[c])];
3e473156
OK
2301 } else {
2302 // latin-1 (direct)
38444daf 2303 for (size_t c=0; c<n; c++)
4e57b0d4 2304 buf[c] = (unsigned char)(psz[c]);
3e473156 2305 }
38444daf 2306 return n;
2bb67b80 2307 }
38444daf 2308 return strlen(psz);
2bb67b80
OK
2309}
2310
6b769f3d 2311size_t wxCSConv::WC2MB(char *buf, const wchar_t *psz, size_t n) const
2bb67b80 2312{
a28d23bb 2313 ((wxCSConv *)this)->LoadNow(); // discard constness
3e473156 2314 if (buf) {
a28d23bb 2315 if (m_cset) {
38444daf 2316 for (size_t c=0; c<n; c++) {
4e57b0d4
VZ
2317 size_t n;
2318 for (n=0; (n<256) && (m_cset->data[n] != psz[c]); n++);
2319 buf[c] = (n>0xff) ? '?' : n;
3e473156
OK
2320 }
2321 } else {
2322 // latin-1 (direct)
38444daf 2323 for (size_t c=0; c<n; c++)
4e57b0d4 2324 buf[c] = (psz[c]>0xff) ? '?' : psz[c];
3e473156 2325 }
38444daf 2326 return n;
2bb67b80 2327 }
38444daf 2328 return wcslen(psz);
2bb67b80 2329}
eea4f86a
OK
2330
2331#endif//wxUSE_WCHAR_T
0f3e3e0c
JS
2332
2333#if wxUSE_WCHAR_T
2334const wxWCharBuffer wxMBConv::cMB2WC(const char *psz) const
2335 {
2336 if (psz) {
2337 size_t nLen = MB2WC((wchar_t *) NULL, psz, 0);
2338 wxWCharBuffer buf(nLen);
2339 MB2WC(WCSTRINGCAST buf, psz, nLen);
2340 return buf;
2341 } else return wxWCharBuffer((wchar_t *) NULL);
2342 }
2343
2344const wxCharBuffer wxMBConv::cWC2MB(const wchar_t *psz) const
2345 {
2346 if (psz) {
2347 size_t nLen = WC2MB((char *) NULL, psz, 0);
2348 wxCharBuffer buf(nLen);
2349 WC2MB(MBSTRINGCAST buf, psz, nLen);
2350 return buf;
2351 } else return wxCharBuffer((char *) NULL);
2352 }
2353
2354#endif//wxUSE_WCHAR_T
2355